%0 Journal Article
%T Distal and proximal family predictors of adolescents' smoking initiation and development: A longitudinal latent curve model analysis
%A Tore Tjora
%A J£¿rn Hetland
%A Leif Aar£¿
%A Simon £¿verland
%J BMC Public Health
%D 2011
%I BioMed Central
%R 10.1186/1471-2458-11-911
%X We employed data from The Norwegian Longitudinal Health Behaviour Study (NLHB), in which participants were followed from the age of 13 to 30. We analysed data from the first 5 waves, covering the age span from 13 to 18, with latent curve modeling (LCM).Smoking rates increased from 3% to 31% from age 13 to age 18. Participants' smoking was strongly associated with their best friends' smoking. Parental SES, parents' smoking and older siblings' smoking predicted adolescents' initial level of smoking. Furthermore, the same variables predicted the development of smoking behaviour from age 13 to 18. Parents' and siblings' smoking behaviours acted as mediators of parents' SES on the smoking habits of adolescents.Parents' SES was significantly associated, directly and indirectly, with both smoking initiation and development. Parental and older siblings' smoking behaviours were positively associated with both initiation and development of smoking behaviour in adolescents. There were no significant gender differences in these associations.Among adolescents in the US and elsewhere, smoking rates have declined [1,2]. However, despite a general decline over cohorts and the well-documented and well-disseminated fact that smoking has severe health consequences, many adolescents still start smoking.It therefore remains an important question what leads young non-smokers to take up smoking. Candidate factors such as the smoking behaviours of parents, siblings and friends are positively correlated with adolescents' smoking [3-6]. Micro-level (proximal) factors such as perceived social pressure from peers, parents and siblings regarding smoking can form subjective norms that may either increase or reduce the chance that adolescents will start smoking. Smoking may also be influenced directly or indirectly by macro-level (distal) factors such as parents' socioeconomic status (SES) [7].For adult populations, the association between SES and smoking is well established [8-10]: people with a
%U http://www.biomedcentral.com/1471-2458/11/911